|Year : 2021 | Volume
| Issue : 2 | Page : 33-39
Consensus molecular subtyping of colorectal cancer by immunohistochemistry, an imperative for a resource limited setting: Report of a Nigerian study
Fatimah B Abdulkareem1, Galina Khramtsova2, Lateef A Odukoya3, Kabir B Badmos1, Tunde Adedokun4, Olorunda Rotimi5, Abiola Ibraheem6, Andrey Khramtsov7, Lise Sveen2, Ian Hurley2, Masaya Hattori2, Dezheng Huo2, Olufunmilayo I Olopade2
1 Department of Anatomic and Molecular Pathology, College of Medicine, University of Lagos & Lagos University Teaching Hospital, Idi-Araba, Nigeria
2 Center for Clinical Cancer Genetics and Global Health, Department of Medicine, University of Chicago, Chicago, IL, USA
3 Department of Anatomic and Molecular Pathology, Lagos University Teaching Hospital, Idi-Araba, Lagos, Nigeria
4 University of Chicago, Chicago, IL, USA
5 Department of Cellular Pathology, University of Leeds Teaching Hospital, Leeds, UK
6 Morehouse School of Medicine, Atlanta, GA, USA
7 Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL, USA
|Date of Submission||10-Jan-2022|
|Date of Acceptance||13-Jan-2022|
|Date of Web Publication||24-Feb-2022|
Prof. Fatimah B Abdulkareem
Department of Anatomic and Molecular Pathology, College of Medicine University of Lagos, Idi-Araba.
Source of Support: None, Conflict of Interest: None
Background and Objectives: Studies of colorectal cancer (CRC) molecular heterogeneity have used genome-wide gene expression-based data to group patients into four consensus molecular subtypes (CMS), but the cost and sophistication of analysis has limited its clinical application. This study aimed at using immunohistochemistry (IHC) to classify CRC specimens in a cohort of patients in Lagos University Teaching Hospital. Materials and Methods: Tissue microarrays were constructed from 75 FFPE tissue blocks of CRC. These were stained for mismatch repair (MMR) proteins (MLH1, MSH2, MSH6, and PMS2) and four other markers (CDX2, HTR2B, ZEB1, and Ki-6) by IHC. Semi-quantitative scoring was performed for the other four markers. A panel of CDX2, HTR2B, and ZEB1 was then used to distinguish between CMS4 and CMS2/CMS3 subtypes, whereas Ki-67 was used to separate CMS2 from CMS3 subtype. MMR status was used to identify CMS1 subtype. Results: Of the total evaluable 75 CRC cases, 38% were <40 years old, 60% were males, with mean of 44.8 years (standard deviation [SD] = 16.1). Fifty-nine patients (79%) had microsatellite stable (MSS) tumor, and the remaining 16 (21%) had microsatellite unstable (MSI) tumor (i.e., CMS1). Thirty-seven (49%) were classified as CMS2 (n = 24) or CMS3 (n = 13) and 22 (29%) of the cases were classified as CMS4. The CMS4 subtype was significantly more likely to occur among young patients (P < 0.001). CMS1 subtype was more in patients older than 40 years and 75% of right-sided cancers were CMS1 (P < 0.001). Conclusion: The study confirms that IHC-based CMS classification and stratification of CRC patients could be a cost-effective prognostic and predictive tool suitable for resource-limited settings.
Keywords: CMS, colorectal, immunohistochemistry, molecular, subtyping
|How to cite this article:|
Abdulkareem FB, Khramtsova G, Odukoya LA, Badmos KB, Adedokun T, Rotimi O, Ibraheem A, Khramtsov A, Sveen L, Hurley I, Hattori M, Huo D, Olopade OI. Consensus molecular subtyping of colorectal cancer by immunohistochemistry, an imperative for a resource limited setting: Report of a Nigerian study. Niger J Gastroenterol Hepatol 2021;13:33-9
|How to cite this URL:|
Abdulkareem FB, Khramtsova G, Odukoya LA, Badmos KB, Adedokun T, Rotimi O, Ibraheem A, Khramtsov A, Sveen L, Hurley I, Hattori M, Huo D, Olopade OI. Consensus molecular subtyping of colorectal cancer by immunohistochemistry, an imperative for a resource limited setting: Report of a Nigerian study. Niger J Gastroenterol Hepatol [serial online] 2021 [cited 2022 Jun 30];13:33-9. Available from: https://www.njghonweb.org/text.asp?2021/13/2/33/338249
| Introduction|| |
Colorectal cancer (CRC) is the third most common cancer and the second most frequent cause of cancer-related deaths worldwide. The incidence of CRC is rising in Nigeria, where it now ranks the fourth most common cancer and equally contributes to a significant proportion of cancer-related mortality. In addition, an increasing number of CRC in young patients has been reported.,
Douaiher et al. projected that the global incidence of CRC would increase by 80% in 2035 with a significant rise in the young and underdeveloped countries. Irabor et al. reported a threefold increase in incidence from the Ibadan cancer registry data to an average of 70 cases per annum between 2002 and 2006. In a systematic review of 2497 histologically diagnosed CRC cases reported in the Nigerian literature over 53 years (1954–2017), we reported an increase in average number of cases published per annum from 18.3 to 86.8. The mean ages in studies analyzed was 46.2 years 32.2% of cases were below 40 years.
CRC is heterogeneous with multiple genetic and epigenetic alterations underlying its pathogenesis. The different subtypes have distinct morphological and molecular characteristics which explain the differences in disease outcomes and response to therapy. This heterogeneity hindered a clinically relevant classification but has motivated the efforts toward the search for a molecular classification that best categorizes these tumors into clinically relevant and prognostically significant subtypes. One of the earlier efforts classified CRC into chromosomal instability (CIN), hypermutated (microsatellite instability [MSI]), and serrated pathway. The CIN tumors consisted of left-sided tumors that were associated with oncogene activation (KRAS, PIK3CA) and tumor suppressor gene inactivation (e.g., APC, SMAD4, TP53, WNT), MSI tumors had defective DNA mismatch repair (MMR), whereas serrated pathway tumors had overlapping features of CIN and MSI.,
With the advent of Tumor Cancer Genome Atlas More Details, TCGA, molecular classification using array-based and sequencing technologies using genomic and transcriptomic characterization was proposed in 2013. TCGA defined three subtypes: hypermutated (13%), ultramutated (3%), and CIN (84%). Application of this classification approach was associated with inconsistencies; thus, an international consortium of six expert groups developed another classification, the CMS subtyping which was reported to be more robust and has better biological interpretation and clinical relevance. This classification is based on gene expression profiling of CRCs after coalescing information from CRC gene expression datasets., Datasets from 18 CRC gene expression studies were used to classify CRCs into four molecular subtypes, which are believed to carry clinical and prognostic significance for patient management.
First among these subtypes is CMS1 (MSI-immune; 14%), these are hypermutated tumors with microsatellite instability and strong immune activation (PD1 activation, NK cell, Th1 cell, and cytotoxic T cell infiltration signatures). They frequently harbor BRAF mutations and have low single copy number alterations (SCNAs). CMS1 patients show a good prognosis but poor survival after relapse. CMS2 (Cannonical, 37%), consists of tumors that show epithelial signatures with increased WNT and MYC signaling activation. They equally harbor loss of tumor suppressor genes and mutations in oncogenes, patients have better survival after relapse. CMS3 subtype (metabolic, 13%), are epithelial tumors with evidence of metabolic dysregulation. They frequently harbor KRAS mutations. CMS4 (Mesenchymal, 23%), tumors show increased expression of EMT genes and prominent transforming growth factor-beta activation, angiogenesis, and stromal invasion. CMS4 tumors display worse overall and relapse-free survival rates. The remaining 13% of cases were designated as mixed type CRCs and these are thought to represent tumors with transition phenotype or intratumoral heterogeneity.,,
Despite the novelty of the CMS classification, there remains the challenge of properly incorporating this approach into clinical practice and patient management, because it requires a complicated genetic testing procedure, and prohibitive cost of reagents and materials. These challenges are more prevalent in a developing country like Nigeria, making the translation of this classification for patient’s management impossible. Recent advances in molecular cancer diagnostics have led to the development of a protocol that combines MSI testing; by PCR or immunohistochemistry (IHC) with an immunohistochemical assay for four other markers and an online tool that is capable of accurately classifying CRC patients into the four major CMS groups., This method, called IHC-CMS classifier uses IHC to detect antibodies against MMR proteins (MLH1 and MSH2) and cases with high-levels MSI considered unstable, are designated CMS1 (MSI). Other cases are assigned either as “epithelial” (CMS2/CMS3) or “mesenchymal” (CMS4) subtypes by staining for the protein products of four genes (i.e., CDX2, FRMD6, HTR2B, and ZEB1). CDX2 is a transcription factor in intestinal epithelial cells, which is expected to be highly expressed in epithelial-like tumors; HTR2B is a serotonin receptor with high expression in mesenchymal-like tumors; FRMD6 is an adaptor protein linking plasma membrane-associated proteins to actin skeleton and is expressed in colon glandular cells and has a higher expression in mesenchymal-like tumors, whereas ZEB1 is an indicator for epithelial–mesenchymal transition (EMT). KER is a pan-cytokeratin marker that is used to normalize the other markers for tumor content. These markers are assessed using semi-quantitative pathologic scoring system that records the percentage of cells stained and the intensity of the immunohistochemical stain. Although the protocol has a limitation of not being able to clearly discriminate between CMS2 and CMS3 tumors, it has the potential of improving the clinical use of the CMS status by promoting its prognostic and predictive usefulness.
Alatise et al. used a combination of IHC and genetic sequencing techniques to perform a genetic/molecular profiling of CRCs among a cohort of Nigerian patients. However, that study was not aimed at subtyping CRCs according to the CMS classification. This study sought to classify CRC into the four main CMS groups using IHC.
| Materials and methods|| |
Archival formalin-fixed paraffin-embedded tissue blocks of 75 patients diagnosed with CRC in the Department of Anatomic and Molecular Pathology of the Lagos University Teaching Hospital (LUTH) were retrieved. The demographic and clinical data of corresponding patients were equally retrieved from their hospital/ laboratory record. The study was conducted in accordance with the Helsinki Declaration and tissue handling was approved by the LUTH Health Research and Ethics Committee. The tissue microarray (TMA) and IHC were carried out at the University of Chicago.
Tissue microarray construction, immunohistochemistry, and immunoreactivity scoring
TMA of colorectal tumor tissue was constructed according to previously described protocols. Briefly, H and E slides of donor FFPE tissue blocks were reviewed on microscopy to mark representative tumor areas by FBA, KG, and OLA. Tissue cylinders with a 0.6-mm diameter were punched from representative tumor areas of each donor FFPE tissue block and transferred into one recipient paraffin block (30 mm × 25 mm). Each of the TMA spots included at least 50% tumor cells. Experienced pathologists confirmed the histologic tumor types.
IHC was performed as per standard protocol. TMA slides were incubated overnight at 57°C, deparaffinized in xylene, rehydrated through graded of ethanol solution (100%, 90%, 90%, and 70%), quenched for endogenous peroxidase activity in 0.3% hydrogen peroxide at 37°C for 10 min, and heat-induced epitope retrieval. Normal colonic epithelium adjacent to the tumor, inflammatory, lymphoid, stromal, and endothelial cells served as internal positive controls for MMR proteins [Figure 1]. Appropriate external positive (human tonsil, CRC) and negative (tumors known to lack of MMR protein expression) controls were used. Negative controls were isotypic IgG or no primary antibody.
|Figure 1: Immunohistochemical staining of colorectal cancer with MMR markers. Magnification x20. (A) Negative immunohistochemical staining of colorectal cancer with MLH1. Lymphocytes serve as internal positive control. (B) Positive staining with MLH1. (C) Negative staining with MSH2. (D) Positive staining with MLH1|
Click here to view
MMR status was detected by staining with antibodies against MLH1 (Pierce), MSH2 (Life Technologies), MSH6 (Novex), and PMS2 (Pierce). MMR IHC-scoring was performed as described by Koopman et al. Three markers were selected from transcriptomic analysis for the IHC-CMSs classifier, including CDX2 (Life Technologies), HTR2B (Life Technologies), and ZEB1 (eBioScience). Ki-67 (DAKO) was added to the panel as an equivalent marker to beta-catenin activation. Staining was performed following standard protocol as described above and antibodies to the named markers were used. Four observers (FBA, KA, KG, OLA) performed quantitative analysis of the tissue specimen without knowledge of specimen identification. All discrepancies were resolved by a second examination using a multihead microscope. Each sample was scored for both intensity and percentage of cells stained for each of CDX2, HTR2B, ZEB1 and Ki-67. For CDX2, ZEB1 and Ki-67, intensity was scored semi-quantitatively as 0 = absent, 1 = weak, 2 = moderate, and 3 = strong, whereas proportion was assessed as the percentage of cells with positive nuclear stain as against cells with negative stain in a section. For HTR2B which is a cytosolic/membranous stain, the same scoring method was applied. The positive stain was classified as presence of expression, and negative as loss of MMR protein expression. IHC assays were performed using a Dako immunostainer, and the antibodies, dilutions, and antigen retrieval methods used are summarized in [Table 1].
|Table 1: Antibodies and conditions used for immunohistochemical analyses|
Click here to view
Analysis was performed using STATA software 2014 version 14. Descriptive summary statistics such as mean and standard deviation were computed for numerical variables, whereas frequencies and proportions were estimated for categorical variables. Associations between CMS categories and categorical demographic and tumor characteristics were analyzed using Fisher’s exact test, whereas HTR2B and Ki67 were compared between CMS groups using the Fisher’s exact test. A value of P ≤ 0.05 was considered statistically significant.
| Results|| |
Characteristics of cases according to CMS status
Seventy-five cases of CRC were enrolled. The mean age was 44.8 years with a range of 16 to 80 years (standard deviation [SD] = 16.1). Over one-third (38%) of the cases were younger than 40 years of age. Of the 75 cases, 60% (45/75) were male and 40% (30/75) were female. Sixty percent (45/75) of the cases were well-differentiated adenocarcinoma, 22.7% (17/75) were moderately differentiated, poorly differentiated tumors accounted for 8% (6/75), mucinous carcinoma 4%, undifferentiated and signet ring cell carcinomas each constituted 2.7% [Table 2]. Overall, 78.7% (59) of the cases were microsatellite stable (MSS), whereas 21.3% were microsatellite unstable (MSI). Based on this and immunohistochemical staining patterns of the selected markers, 21.3% (16/75) of the cases were CMS1 subtype, 32% (24/75) were categorized as CMS2, 17.3% (13/75) were CMS3 and 29.3% (22/75) as CMS4 subtype [Figure 2].
|Table 2: Clinicopathologic parameters and CMS status of colorectal cancer cases*|
Click here to view
|Figure 2: Frequencies of consensus molecular subtypes of colorectal cancer|
Click here to view
There was a significant association between CMS subtypes and tumor site (P = 0.0001) and age (P = 0.0001). In terms of age, 92.9% of individuals with CMS1 tumors were 40 years and above, whereas 13.6% of those in CMS 4 were 40 years and above (P = 0.0001). The corresponding proportions for CMS 2 and CMS 3 tumors were 72.7% and 92.3%, respectively [Table 2]. CMS 1 tumors were mostly right sided, constituting 80% compared to 8.3% of CMS2, none of CMS3, and 13.6% of CMS 4 tumors (P = 0.0001).
HTR2B was universally expressed; however, a higher median HTR2B score of 3 was observed in CMS4 cases compared to a median score of 1 for the other subtypes (P = 0.0001). Of the fifteen cases with an HTR2B score of 3, 86.7% (13/15) were CMS4 subtype. Ki67 was also universally expressed; however, a higher median score of 2 or 3 was observed in CMS2 cases compared to the other subtypes (P = 0.0001). Of the twenty-three cases with a Ki67 score of 3, 52.2% (12/23) were CMS2 subtype and of the cases with a score of 2 CMS 2 alone constitute 37% (10/27).
| Discussion|| |
The consensus molecular subtyping of CRC currently represents the best attempt at a clinically relevant molecular classification of CRCs as it captures the highly heterogeneous nature of this group of neoplasms. The classification is based on the integration of six independent transcriptomic derived study reports and CMS status has over the past few years become a validated prognostic tool in the diagnosis and management of CRC. Despite the novelty of this method, it has not been easy to adopt in routine pathology practice because the original classification is based on gene expression profiling, which is an expensive, time-consuming, and sophisticated laboratory procedure. IHC-based CMS classifiers were recently developed to bring this molecular classification to bear on routine diagnosis, management, and prognostication of CRCs. A good measure of concordance has been reported between the transcriptome-based profiling of CRC and the IHC-based profiling for the purpose of CMS classification. Trinh et al., for instance, reported a concordance rate of up to 87%. This is a welcome development especially because studies have indicated that CMS subtype is an independent prognostic factor in individuals presenting with metastatic CRC and who are administered first-line therapy. In addition, CMS is a potential marker that can guide the selection of patients likely to benefit from anti-VEGF and anti-EGFR therapy., Owing to its observed benefit inpatient management, our study sought to use IHC in determining the CMS status of CRC diagnosed in Lagos.
We identified 21% of CRC cases as CMS1, 32% as CMS2, 17% as CMS3, and 29% as CMS4, respectively. These proportions concur with reports of the international consortium of experts that used large-scale data from the transcriptome-based gene profiling technique except for CMS1, which according to the traditional classification accounted for about 14% of cases but we found a proportion of 21% (comparison shown in [Figure 3]). This may suggest that more of our patients belong to the microsatellite instability (MSI) category as previously observed by earlier workers.,, Studies have indicated that MSI is elevated in the black populations. Ashktorab et al. recorded 43% proportion of MSI tumors in African Americans compared to <20% in the general population in 2005. However, a meta-analysis of 22 studies within the USA about 10 years after, reported the overall rate of MSI in all the studies analyzed was 17%. Several small IHC-based study cohorts from Nigeria showed rates ranging from 23% to 53%, whereas the Ghanaian study which is based on genetic testing of 10 markers reported 43% [Table 3].,[25-27] A more recent study on molecular and phenotypic profiling of CRC in West Africa by Alatise et al. reported 28.1% of the 64 Nigerian specimens that underwent MSK-IMPACT, to be MSI-high and 21.3% (20 of 94) by IHC, compared to 7.2% (7 of 97) African American in the cases from MSKCC. All of these reported findings strongly suggest that a significant proportion in Nigerian CRCs might fall into the CMS 1 category [Table 1].
|Figure 3: Colorectal cancer molecular subtypes by immunohistochemistry in a patient-cohort of 75 cases from Nigeria compared to Guiney et al.|
Click here to view
|Table 3: FrequenciesMSI-positive colorectal cancers detected by different techniques reported in different studies in Nigeria and Ghana|
Click here to view
CMS1 subtype of CRCs has a predilection for the proximal colon, whereas CMS2 has a predilection for the distal colon or rectum according to observations in other studies. In our study there was a significant association between tumor site and CMS status, with 71% (12/17) of all right-sided tumors identified as CMS1. This agrees with an earlier report by Alatise et al., which found that 66.7% of all MSI-H cases were located on the right colon among a cohort of Nigerian cases they investigated. CMS1 cases equate to the MSI-positive cases, which according to extant characterization are known to mostly have a proximal colon location, and harbor a lot of tumor infiltrating lymphocytes. Most of the CRC involving the left side were CMS2, about 39% (22/57)and distal location equally predominates for CMS3 and CMS4. This closely mirrors the report of the recent study by Alatise et al., which observed a high frequency of MSS tumors in left-sided CRC cases. Overall, left-sided CRC are the most common and our study concurs with this finding as 77% of the cases investigated involved the distal colon.
Increasing incidence of CRC has been reported among young patients in Nigeria; and in line with this trend, we observed that greater than one-third of the cases we evaluated were younger than 40 years. Sixty-three percent of the samples investigated were from persons younger than 50 years old. Previous studies from Nigeria have reported a similar trend.,, Earlier, we had found 32.2% of 2497 cases in a systematic review of CRCs in Nigeria to occur under 40 years. The mean ages in most studies ranged from 39 to 50.7 years with an average of 46.2 years. Similar trends have been observed in Western countries instigating a renewed interests in investigating factors responsible for the rise in early-onset CRC with a view to lowering the age of screening initiation to 45 years. Putative factors suggested to be responsible for this rise include changing diet, obesity, and other lifestyle factors.
This study found that CMS subtype significantly varied when compared with patients’ age distribution. Approximately 93% of CMS1 cases were aged 40 years and above and 86% of cases designated CMS4 were younger than 40 years. This finding agrees with earlier works that have described CMS1 as adult-onset cancer. However, it contrasts with those of Willauer et al., who reported a higher proportion of CMS1 in patients younger than forty and observed that CMS3 and CMS4 were less common below 40 years.
The CMS statuses in this study showed no significant association with gender, and the histological type of CRCs. CMS1 has been suggested to be more prevalent in females. However, no significant differences between the genders were observed in this study. Alatise et al. also reported no significance differences with regards to age and gender of patients with MSI-H tumors in Nigeria. Similarly, Li et al. in their study of 165 CRCs equally reported no significant relationship between CMS status and gender, as well as CMS status and histological type or degree of tumor differentiation. Therefore, no differences exist between the CMS subtypes and tumor differentiation. However, this needs to be further examined in larger future studies.
The current study slightly differs in approach in terms of the markers used from that by Trinh et al. We distinguished between CMS2 and CMS3 by using Ki67 to discriminate between the two epithelial like subtypes. This is an improvement on the limitation of similar studies conducted by other authors in the past where CMS 2 and CMS 3 were not separated. Although CMS2 and CMS3 are known to have similar prognoses, it is important to separate the subtypes because of differences that exist due to a major metabolic dysregulation reportedly associated with CMS3.
An important limitation of the study is a small sample size, which could have caused some statistical distortion. Regardless, our study shows the possibility of a CMS classification of CRCs, using IHC, which will be a cost-effective way to prognosticate and select patients for the purpose of treatment in the face of scarce resources.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al
. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2021;71:209-49.
Global Cancer Observatory Population Fact Sheets. Nigeria: Summary Statistics. 2020;2019–9. Available from: https://gco.iarc.fr/today/data/factsheets/populations/566-nigeria-fact-sheets.pdf. [Last accessed on 2021 Dec 14].
Abdulkareem FB, Abudu EK, Awolola NA, Elesha SO, Rotimi O, Akinde OR, et al
. Colorectal carcinoma in Lagos and Sagamu, southwest Nigeria: A histopathological review. World J Gastroenterol 2008;14:6531-5.
Irabor DO, Afuwape OO, Ayandipo OO The present status of the management of colon and rectal cancer in Nigeria. 2014;2014:267190.
Douaiher J, Ravipati A, Grams B, Chowdhury S, Alatise O, Are C Colorectal cancer: Global burden, trends, and geographical Variations. Surg Oncol 2017;115:619-630.
Irabor D, Adedeji OA Colorectal cancer in Nigeria: 40 years on: A review. Eur J Cancer Care (Engl) 2009;18:110-5.
Rotimi O, Abdulkareem FB Fifty-three years of reporting colorectal cancer in Nigerians: A systematic review of the published literature. Niger Postgrad Med J 2014;21:68-73.
Lee GH, Malietzis G, Askari A, Bernardo D, Al-Hassi HO, Clark SK Is right-sided colon cancer different to left-sided colorectal cancer? Eur J Surg Oncol 2014;41:300-8.
Baran B, Ozupek NM, Tetik NY, Acar E, Bekcioglu O, Baskin Y et al
. Differences btw right sided and left sided colorectal cancer. Gastroenterol Res 2018;11:264-73.
Muller MF, Ibrahim AEK, Arends MJ Molecular pathological classification of colorectal cancer. Virchows Arch 2016;469:125-34.
Guinney J, Dienstmann R, Wang X, de Reyniès A, Schlicker A, Soneson C, et al
. The consensus molecular subtypes of colorectal cancer. Nat Med 2015;21:1350-6.
Davis LE The evolution of biomarkers to guide the treatment of metastatic colorectal cancer. Am J Manag Care 2018;24:107-17.
Li X, Larsson P, Ljuslinder I, Ling A, Löfgren-Burström A, Zingmark C, et al
. A modified protein marker panel to identify four consensus molecular subtypes in colorectal cancer using immunohistochemistry. Pathol Res Pract 2021;220:153379.
Li Y, Yao Q, Zhang L, Mo S, Cai S, Huang D, et al
. Immunohistochemistry-based consensus molecular subtypes as a prognostic and predictive biomarker for adjuvant chemotherapy in patients with stage II colorectal cancer. Oncologist 2020;25:e1968-79.
Alatise OI, Knapp GC, Sharma A, Chatila WK, Arowolo OA, Olasehinde O, et al
. Molecular and phenotypic profiling of colorectal cancer patients in West Africa reveals biological insights. Nat Commun 2021;12:6821.
Sauter G, Simon R, Hillan K Tissue microarrays in drug discovery. Nat Rev Drug Discov 2003;2:962-72.
Koopman M, Kortman GAM, Mekenkamp L, Ligtenberg MJL, Hoogerbrugge N, Antonini NF, et al
. Br J Cancer 2009;100:266-73.
Trinh A, Trumpi K, De Sousa E Melo F, Wang X, de Jong JH, Fessler E, et al
. Practical and robust identification of molecular subtypes in colorectal cancer by immunohistochemistry. Clin Cancer Res 2017;23:387-98.
Okita A, Takahashi S, Ouchi K, Inoue M, Watanabe M, Endo M, et al
. Consensus molecular subtypes classification of colorectal cancer as a predictive factor for chemotherapeutic efficacy against metastatic colorectal cancer. Oncotarget 2018;9:18698-711.
Lenz HJ, Ou FS, Venook AP, Hochster HS, Niedzwiecki D, Goldberg RM, et al
. Impact of consensus molecular subtype on survival in patients with metastatic colorectal cancer: Results from CALGB/SWOG 80405 (Alliance). J Clin Oncol 2019;37:1876-85.
Irabor DO, Oluwasola OA, Ogunbiyi OJ, Ogun OG, Okolo CA, Melas M, et al
. Microsatellite instability is common in colorectal cancer in native Nigerians. Anticancer Res 2017;37:2649-54.
Aminu DMC, Yawale I, Ibrahim Y Microsatellite instability profile of colorectal cancers in Kano Northwestern Nigeria. Arch Int Surg 2017;7:126-31.
Ashktorab H, Smoot DT, Farzanmehr H, Fidelia-Lambert M, Momen B, Hylind L, et al
. Clinicopathological features and microsatellite instability (MSI) in colorectal cancers from African Americans. Int J Cancer 2005;116:914-9.
Ashktorab H, Ahuja S, Kannan L, Llor X, Ellis NA, Xicola RM, et al
. A meta-analysis of MSI frequency and race in colorectal cancer. Oncotarget 2016;7:34546-57.
Duduyemi BM, Akang EEU, Adegboyega PA, Thomas JO Significance of DNA mismatch repair genes and microsatellite instability in colorectal carcinoma in Ibadan, Nigeria. Am J Med Bio Res 2013;1:145-8.
Adegoke OO, Komolafe AO, Ojo OS Microsatellite instability statuses and clinicopathological characteristics of colorectal carcinomas in a sub-Saharan African population. Gastroenterol Hepatol Int J 2017;2:000119.
Raskin L, Dakubo JC, Palaski N, Greenson JK, Gruber SB Distinct molecular features of colorectal cancer in Ghana. Cancer Epidemiol 2013;37:556-61.
Inamura K Colorectal cancers: An update on their molecular pathology. Cancers (Basel) 2018;10:26.
Willauer AN, Liu Y, Pereira AAL, Lam M, Morris JS, Raghav KPS, et al
. Clinical and molecular characterization of early-onset colorectal cancer. Cancer 2019;125:2002-10.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3]